Advances in genomic and molecular technologies have generated great interest in the concept of personalized patient care, particularly in oncology where heterogeneity of both toxicity and response has been reported. Some of this variability can be predicted on the basis of clinical and genomic patient characteristics. Development of clinical and genomic screens would enable prospective identification of patients at high risk for toxicity, allowing use of screening and preventative measures to improve management. Using data and specimens from the RTOG 0825, specific aims to be addressed include: (1) Screen and prioritize clinical risk factors and genetic variants associated with temozolomide-related myelotoxicity and build and cross-validate risk prediction models for toxicity that will be incorporated into a web-based patient management tool; (2) Screen and prioritize clinical risk factors and genetic variants associated with bevacizumab-related vascular toxicity and build and cross-validate risk prediction models for toxicity that will be incorporated into a web- based patient management tool; (3) Determine the: a) concordance between presence of single nucleotide polymorphisms (SNPs) in genomic DNA from white blood cells and tumor tissue; and b) the relationship of treatment toxicity and toxicity-related SNPs on outcomes (progression-free and overall survival) after cytotoxic (temozolomide) and antiangiogenic (bevacizumab) therapy. Our proposal uniquely enables us to develop a two-stage model that will, ultimately, utilize clinical factors to screen for patients at increased ris and then perform genetic testing only on those select patients to refine the risk assessment, thereby improving clinical utility and feasibility. Clinical data and biospecimens (blood and tumor) collected during the RTOG 0825 clinical trial will be used to develop clinical risk models supplemented with genomic variants (single nucleotide polymorphisms-SNPs) related to myelotoxicity due to temozolomide, vascular toxicities due to bevacizumab, and response to therapy (overall and progression-free survival) in patients treated for newly diagnosed GBM. Furthermore, this paradigm combining clinical factors with genomics is designed to evaluate the relationship between toxicity and tumor response, thereby achieving the goal of establishing a modified therapeutic window. These paradigms will be developed into a useful clinical computer application to interpret the results of the final models for individual patients. The project is highly translational in that the findings of this study have strong potential for personalized care by identifying patient-specific variability in treatment toxicity and response. The investigative paradigm leverages the clinical data and samples already collected as part of a recently completed large cooperative group trial. In addition, the paradigm of creating a clinically-based assessment that can better define the patients at risk will improve the utility, feasibility, and cost-benefit of genomic screening that has the potential application to other soli tumors and therapies.